A 3D recording device (1) is provided including an image recording device (2), a distance measuring device (3), and an output unit (5). An image processing device (4) is used to calculate, for a sequence (8) of images (9, 10) recorded in different poses (15, 16, 17), groups (29, 30, 31, 32) of image elements (18, 19, 20, 21, 25, 26, 27, 28) corresponding to each other and to determine for each group (29, 30, 31, 32) a three-dimensional position indication (48, 49, 50) and to scale the three-dimensional position indication (48, 49, 50) with the aid of distance information (42) measured by the distance measuring device (3).
Legal claims defining the scope of protection, as filed with the USPTO.
1. A 3D recording device ( 1 ) comprising: an image-recording device ( 2 ), a rangefinder ( 3 ) and an image-processing device ( 4 ), the image-processing device ( 4 ) is configured to: identify ( 33 ) at least one group ( 29 , 30 , 31 , 32 ) of mutually corresponding image constituents ( 18 , 19 , 20 , 21 , 25 , 26 , 27 , 28 ) in a sequence ( 8 ) of at least two images ( 9 , 10 ) recorded by the image-recording device ( 2 ), calculate ( 38 ) at least one three-dimensional position specification ( 50 ) in relation to the at least one group ( 29 , 30 , 31 , 32 ) of the mutually corresponding image constituents ( 18 , 19 , 20 , 21 , 25 , 26 , 27 , 28 ) from a respective image position ( 59 , 60 ) of the image constituents ( 18 , 19 , 20 , 21 , 25 , 26 , 27 , 28 ) from the group ( 29 , 30 , 31 , 32 ) in the images ( 9 , 10 ) of the sequence ( 8 ), calculate ( 38 ) a recording pose of the image-recording device ( 2 ) from the respective image position ( 59 , 60 ) of each of the image constituents ( 18 , 19 , 20 , 21 , 25 , 26 , 27 , 28 ) of the at least one group ( 29 , 30 , 31 , 32 ), and then scale ( 46 ) a calculated at least one three-dimensional position specification ( 50 ) using measured distance information ( 42 ) that is measured by the rangefinder ( 3 ); and calculate further three-dimensional distance information associated with the at least one group ( 29 , 30 , 31 , 32 ) of mutually corresponding image constituents ( 18 , 19 , 20 , 21 , 25 , 26 , 27 , 28 ) for which no distance measurement is available, in at least one of an extrapolation or interpolation method ( 47 ) from the measured distance information ( 42 ) and the calculated at least one three-dimensional position specification ( 50 ).
2. The 3D recording device ( 1 ) as claimed in claim 1 , further comprising an output unit ( 5 ) configured to output the at least one three-dimensional position specification which is scaled.
3. The 3D recording device ( 1 ) as claimed in claim 1 , wherein the image-processing device is configured to calculate further three-dimensional position specifications ( 48 , 49 ) from the calculated three-dimensional position specifications ( 50 ) and recording poses ( 15 , 16 , 17 ).
4. The 3D recording device ( 1 ) as claimed in claim 1 , wherein the image-processing device ( 4 ) is configured to identify ( 33 ) the at least one group ( 29 , 30 , 31 , 32 ) of mutually corresponding image constituents ( 18 , 19 , 20 , 21 , 25 , 26 , 27 , 28 ) using a feature analysis ( 23 ), or the image-processing device ( 4 ) is configured to segment ( 55 ) the images ( 9 , 10 ) into image segments, or the image-processing device ( 4 ) is configured to identify ( 33 ) the at least one group ( 29 , 30 , 31 , 32 ) of mutually corresponding image constituents ( 18 , 19 , 20 , 21 , 25 , 26 , 27 , 28 ) using a feature analysis ( 23 ) and to segment ( 55 ) the images ( 9 , 10 ) into image segments.
5. The 3D recording device ( 1 ) as claimed in claim 1 , wherein an actuator ( 57 ) is embodied to align a measurement direction of the rangefinder ( 3 ) in such a way that the measured distance information ( 42 ) is related to an image segment.
6. The 3D recording device ( 1 ) as claimed in claim 1 , wherein at least one of a movement or acceleration sensor ( 58 ) is configured to detect a movement of the image-recording device ( 2 ).
7. The 3D recording device ( 1 ) as claimed in claim 1 , wherein the image-processing device ( 4 ) is configured to fit a geometric object ( 51 ) into the at least one scaled three-dimensional position specification.
8. The 3D recording device ( 1 ) as claimed in claim 1 , wherein the image-processing device ( 4 ) is configured to calculate at least one of a length, area or volume from the at least one three-dimensional position specification.
9. The 3D recording device ( 1 ) as claimed in claim 1 , wherein the rangefinder ( 3 ) is embodied as a laser rangefinder.
10. The 3D recording device ( 1 ) as claimed in claim 1 , wherein the image-processing device is configured to calculate further three-dimensional position specifications ( 48 , 49 ) from the calculated three-dimensional position specifications ( 50 ) and recording poses ( 15 , 16 , 17 ) using a stereo method or from an interface calculated by a level set method.
11. The 3D recording device ( 1 ) as claimed in claim 1 , wherein the rangefinder ( 3 ) is configured to modify an alignment of at least one of a measurement direction or a measurement beam ( 52 ).
12. The 3D recording device ( 1 ) as claimed in claim 1 , wherein the rangefinder ( 3 ) is embodied to generate a measurement beam ( 52 ) in a spectral range detectable by the image-recording device ( 2 ).
13. The 3D recording device ( 1 ) as claimed in claim 1 , wherein the image-processing device ( 4 ) is configured to calculate a three-dimensional representation of the at least one group ( 29 , 30 , 31 , 32 ) of the mutually corresponding image constituents ( 18 , 19 , 20 , 21 , 25 , 26 , 27 , 28 ) at the at least one scaled three-dimensional position specification.
14. The 3D recording device ( 1 ) as claimed in claim 1 , wherein the output unit ( 5 ) is configured to graphically output the at least one group ( 29 , 30 , 31 , 32 ) of mutually corresponding image constituents ( 18 , 19 , 20 , 21 , 25 , 26 , 27 , 28 ) at an image point corresponding to the three-dimensional position specification ( 48 , 49 , 50 ).
15. A method for producing a 3D image, comprising recording a sequence ( 8 ) of at least two images ( 9 , 10 ) of a scene ( 11 ) using an image-recording device ( 2 ) and identifying at least one group ( 29 , 30 , 31 , 32 ) of mutually corresponding image constituents ( 18 , 19 , 20 , 21 , 25 , 26 , 27 , 28 ) in the recorded images ( 9 , 10 ), calculating at least one three-dimensional position specification ( 50 ) relating to the at least one group ( 29 , 30 , 31 , 32 ) of mutually corresponding image constituents ( 18 , 19 , 20 , 21 , 25 , 26 , 27 , 28 ) in an image-processing device ( 4 ), calculating at least one recording pose ( 15 , 16 , 17 ) of the image-recording device ( 2 ) at a recording time from a respective image position ( 59 , 60 ) of the image constituents ( 18 , 19 , 20 , 21 , 25 , 26 , 27 , 28 ) from the at least one group ( 29 , 30 , 31 , 32 ) in the images ( 9 , 10 ) of the sequence ( 8 ), measuring at least one item of distance information ( 42 ) in relation to the scene ( 11 ) using a rangefinder ( 3 ), and then scaling ( 46 ) the three-dimensional position specification using the measured distance information ( 42 ), and calculating further three-dimensional distance information associated with the at least one group ( 29 , 30 , 31 , 32 ) of mutually corresponding image constituents ( 18 , 19 , 20 , 21 , 25 , 26 , 27 , 28 ) for which no distance measurement is available, in at least one of an extrapolation or interpolation method ( 47 ) from the measured distance information ( 42 ) and the calculated at least one three-dimensional position specification ( 50 ).
16. The method as claimed in claim 15 , further comprising calculating further three-dimensional position specifications ( 48 , 49 ) using a stereo method, or deriving the further three-dimensional position specifications ( 48 , 49 ) from an interface calculated using a level set method.
17. The method as claimed in claim 15 , further comprising identifying the at least one group ( 29 , 30 , 31 , 32 ) of mutually corresponding image constituents ( 18 , 19 , 20 , 21 , 25 , 26 , 27 , 28 ) using a feature analysis ( 23 ) or segmenting the recorded images ( 9 , 10 ) using by the image-processing device ( 4 ).
18. The method as claimed in claim 15 , further comprising measuring the distance information ( 50 ) in relation to a scene constituent ( 34 , 35 , 36 , 37 ), which corresponds to the at least one group ( 29 , 30 , 31 , 32 ) of mutually corresponding image constituents ( 18 , 19 , 20 , 21 , 25 , 26 , 27 , 28 ), automatically.
19. The method as claimed in claim 15 , further comprising calculating the scaling ( 46 ) from the distance information ( 42 ) associated with the at least one group ( 29 , 30 , 31 , 32 ) of mutually corresponding image constituents ( 18 , 19 , 20 , 21 , 25 , 26 , 27 , 28 ), or measuring distance information ( 42 ) in relation to each image segment of the images ( 9 , 10 ).
20. The method as claimed in claim 15 , further comprising measuring the distance information ( 42 ) using a measurement beam ( 52 ), which lies in a spectral range in which the image-recording device ( 2 ) records.
21. The method as claimed in claim 15 , further comprising calculating the at least one three-dimensional position specification ( 48 , 49 , 50 ) using at least one of a SLAM method, or structure-from-motion method.
22. The method as claimed in claim 15 , further comprising identifying a spatial region ( 59 ), in which further three-dimensional position specifications ( 48 , 49 ) are to be calculated, and actuating the rangefinder ( 3 ) in such a way that the distance information ( 42 ) for the spatial region ( 59 ) is measured or a current recording pose ( 17 ) is calculated or estimated from previously calculated recording poses ( 15 , 16 ) for the purposes of actuating the rangefinder ( 3 ).
23. The method as claimed in claim 15 , further comprising fitting a geometric object ( 51 ) into the at least one three-dimensional position specification ( 48 , 49 , 50 ).
24. The method as claimed in claim 15 , further comprising calculating at least one of a length, area or volume from the at least one three-dimensional position specification.
25. A method for setting up a 3D recording device for carrying out the method of claim 15 , comprising connecting the rangefinder ( 3 ) to the image-recording device ( 2 ), measuring the at least one item of distance information ( 42 ) for each of the images ( 9 , 10 ), determining an image position in the associated image ( 9 , 10 ) for each said item of measured distance information ( 42 ), to which image position said distance information ( 42 ) relates, and calculating at least one parameter from the measured distance information ( 42 ) and the determined image positions, said parameter describing an alignment of the rangefinder ( 3 ) relative to the image-recording device ( 2 ).
26. The method as claimed in claim 25 , further comprising projecting at least one of a known or an artificially generated scene with a measurement beam ( 52 ) of the rangefinder ( 3 ).
27. The method as claimed in claim 15 , further comprising calculating further three-dimensional position specifications ( 48 , 49 ) in relation to further image constituents before or after the scaling.
28. The method as claimed in claim 15 , further comprising calculating a movement of the image-recording device ( 2 ) between two individual recordings using an output signal of at least one of a movement or acceleration sensor ( 58 ).
29. The method as claimed in claim 15 , further comprising calculating a three-dimensional representation of the at least one group ( 29 , 30 , 31 , 32 ) of mutually corresponding image constituents ( 18 , 19 , 20 , 21 , 25 , 26 , 27 , 28 ) at the at least one scaled three-dimensional position specification.
30. The method as claimed in claim 15 , further comprising outputting the at least one group ( 29 , 30 , 31 , 32 ) at an image point corresponding to the at least one three-dimensional position specification ( 48 , 49 , 50 ) using an output unit ( 5 ).
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
June 2, 2014
December 18, 2018
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.